Early Detection of Autism through Acoustic Analysis of Cry

通过哭声声学分析早期发现自闭症

• 批准号: 7841504
• 负责人: STEPHEN J SHEINKOPF
• 金额: $ 25.71万
• 依托单位: WOMEN AND INFANTS HOSPITAL-RHODE ISLAND
• 依托单位国家: 美国
• 财政年份: 2009
• 资助国家: 美国
• 起止时间: 2009-09-11 至 2011-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/7841504
• 关键词: Acoustics; Adult; Affective; Age-Months; Algorithms; Area; Arts; Audiotape; Autistic Disorder; Birth; Characteristics; Child; Classification; Clinical; Collaborations; Communication; Communication impairment; Computers; Computing Methodologies; Crying; Data; Databases; Detection; Development; Developmental Delay Disorders; Devices; Diagnosis; Diagnostic; Digital Signal Processing; Disease; Early Diagnosis; Early identification; Emotional; Engineering; Funding; Grant; Home environment; Individual; Infant; Intervention; Lead; Life; Loudness; Measurement; Methods; Modeling; National Institute of Mental Health; National Institute on Deafness and Other Communication Disorders; Operating System; Output; Pattern; Phonation; Plant Roots; Population; Procedures; Process; Production; Property; Public Health; Publishing; Research; Research Personnel; Risk; Screening procedure; Siblings; Signal Transduction; Speech; System; Systems Analysis; Technology; Time; United States National Institutes of Health; Update; Videotape; Visual; Work; analog; autism spectrum disorder; base; computerized data processing; design; digital; early childhood; experience; flexibility; improved; infancy; public health relevance; speech recognition; tool; vocalization

DESCRIPTION (provided by applicant): Preliminary evidence suggests that children with Autism Spectrum Disorders (ASD) may be characterized by atypical features of cry during early infancy. Previous research has identified atypicalities in prosodic qualities of speech and vocalizations in older children with autism, and unusual acoustic features of infant cry would be consistent with such findings from early childhood. Our lab has conducted studies of cry acoustics in a number of clinical populations, including (most recently) infants later diagnosed with an ASD. The cry analysis system utilized in our lab is specifically modeled on the infant's developing vocal tract. However, current acoustic analysis systems, including our own, may not be accurate enough to detect these atypicalities, especially where they may be subtle. This project teams researchers with expertise in infant cry production and autism in infancy and early childhood, as well as a lab with extensive experience in the development of acoustic analysis and speech recognition tools. This team will work collaboratively to develop a new acoustic analysis system for infant cry based on state of the art acoustical engineering and computer technology. In particular, we will utilize advances in digital signal processing to yield more accurate and robust detection and modeling of infant cry acoustics. We will then validate this new system against our current system by using a large data bank of cries already collected, including cries of infants who were later diagnosed with autism spectrum disorder. This project could lead to a simple, non-invasive procedure that could identify signs of autism shortly after birth and would have enormous public health significance. PUBLIC HEALTH RELEVANCE: This research is highly relevant to both early identification and intervention for children with autism. By developing a system that can identify atypical aspects of cry in infancy associated with this disorder, this research will aide in the development of more sensitive and specific screening and diagnostic tools for use in infancy. In addition, this research will improve our understanding of the roots and course of communication impairments in autism.

描述（由申请人提供）：初步证据表明，自闭症谱系障碍（ASD）儿童的特征可能是婴儿早期的非典型哭泣特征。先前的研究已经确定了自闭症大龄儿童言语和发声的韵律质量的相似性，婴儿哭声的异常声学特征与幼儿期的这些发现一致。我们的实验室已经在许多临床人群中进行了哭泣声学的研究，包括（最近）后来被诊断为ASD的婴儿。我们实验室使用的哭声分析系统是专门模仿婴儿发育中的声道的。然而，目前的声学分析系统，包括我们自己的系统，可能不够准确，无法检测到这些复杂性，特别是在它们可能很微妙的地方。该项目团队的研究人员具有婴儿期和幼儿期婴儿哭声产生和自闭症的专业知识，以及在声学分析和语音识别工具开发方面具有丰富经验的实验室。这个团队将合作开发一个新的声学分析系统，婴儿哭的基础上，最先进的声学工程和计算机技术。特别是，我们将利用数字信号处理的进步，以产生更准确和强大的检测和婴儿哭声声学建模。然后，我们将通过使用已经收集的大量哭声数据库来验证这个新系统，包括后来被诊断患有自闭症谱系障碍的婴儿的哭声。该项目可能会导致一种简单的，非侵入性的程序，可以在出生后不久识别自闭症的迹象，并将具有巨大的公共卫生意义。 公共卫生相关性：这项研究与自闭症儿童的早期识别和干预高度相关。通过开发一个系统，可以识别与这种疾病相关的婴儿期哭泣的非典型方面，这项研究将有助于开发更敏感和更具体的筛查和诊断工具，用于婴儿期。此外，这项研究将提高我们对自闭症沟通障碍的根源和过程的理解。